PROJECT SUMMARY Natural visual scenes contain numerous objects, but gaze can be directed to only one at a time. This project has provided numerous insights into the neural mechanisms in the frontal lobe contributing to this essential ability. Research in the next grant period will focus on the frontal eye field (FEF), an area we have shown and others have confirmed contributes to saccade target selection. The overall goal of the next grant period is, in short, to elucidate which neurons in which layers of FEF perform which computations to locate a unique object, resolve the features of that object, determine how to respond to the object, and produce the correct gaze shift toward or away from the object. The goal will be achieved through a sequence of 4 aims. In Aim 1 macaque monkeys will be trained to perform a visual search task in which they must search for an object with a unique color and produce a saccade to an arbitrary location based on the form of that object. The difficulty of the task will be manipulated systematically to resolve the distinct contribution of each sequential visual, cognitive, and motor operation necessary to perform the task correctly. The predominant theory of visual decision making based on a single stage of evidence accumulation following stimulus encoding and preceding saccade production cannot explain performance of this task because it requires a sequence of separate decision operations. Aim 2 will obtain large, dense, unbiased samples of neural signals from all layers of FEF during performance of the task during each causal manipulation. Aim 3 will distinguish functional categories of FEF neurons with unprecedented resolution using a consensus clustering method that we developed. The locations of neural samples will be determined with high-resolution magnetic resonance imaging and verified histologically. This will provide the first high-resolution functional architecture of FEF. Aim 4 will apply a powerful theoretical approach, separate modifiability, to identify which neurons in which layers support which visual, cognitive, and motor operations based on the additive and mutually invariant effects of the distinct task manipulations on neural processing time and accuracy. Information derived from these innovative experiments will establish the most accurate and complete functional and structural description of FEF to date. This will constrain and guide the next generation of computational and microcircuit models of visual search and saccade target selection. Such accurate information is necessary to understand the mechanisms of and develop effective therapies for visual attention and gaze disorders.